A method for array radar target detection based on Fourier transform of space-time image data

The Fourier transform method of space-time graph data is used to separate the targets and clutter in the array radar signal, extract the E-D features and the maximum eigenvalue distance-Doppler diagram, solving the problem of insufficient detection performance of array radar in complex clutter backgrounds. , significantly improves the detection ability of weak targets.

CN118465729BActive Publication Date: 2024-12-06NAVAL AVIATION UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410924220.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-11
Publication Date
2024-12-06
Estimated Expiration
2044-07-11

AI Technical Summary

Technical Problem

The existing array radar has insufficient performance on cross-distance unit and cross-beam motion targets in the context of complex clutter, especially in the detection of weak targets at sea, Doppler feature detection methods have limited performance.

Method used

Using the Fourier transform method based on space-time graph data, the pulse factor graph and array element factor graph are constructed, Cartesian product representation is performed, and Fourier transform is performed to separate the target and clutter signals, and the E-D features and maximum eigenvalues ​​are extracted. Distance-Doppler map for object detection.

Benefits of technology

It improves the detection capability of weak targets in the background of sea clutter, enhances detection performance, and is not affected by the target orientation, and has more stable detection performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118465729B_ABST
    Figure CN118465729B_ABST
Patent Text Reader

Abstract

The present invention relates to an array radar target detection method based on Fourier transform of space-time graph data, and belongs to the technical field of radar target detection methods. In order to solve the problem that the beamforming-moving target detection method in the existing array radar target detection has insufficient detection performance for cross-range unit and cross-beam moving targets under complex clutter background, an array radar target detection method based on Fourier transform of space-time graph data is provided. The method extracts the advantage angle from the high-dimensional complex structure data feature, separates the target and clutter components in the echo signal, and improves the weak target detection capability under the sea clutter background, including the following steps: constructing array radar signal space-time graph data; performing Fourier transform on the space-time graph data; obtaining the E-D feature of the target signal; obtaining the maximum eigenvalue distance-Doppler map of the target signal; and performing target detection on the maximum eigenvalue distance-Doppler map by the existing two-dimensional feature detection method.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to an array radar target detection method, in particular to an array radar target detection method based on Fourier transform of space-time image data, belonging to the technical field of radar target detection methods. Background Art

[0002] Most sea observation radars need to control the beam direction by antenna rotation or electronic scanning to cover a large observation area. However, the radar data update rate is limited by the scanning cycle, the target beam dwell time is short, and the detection capability of low observable targets is limited. In addition, due to the strong ground clutter or sea clutter, the target echo is weak, and the detection performance of CFAR, pulse Doppler processing and other methods is low.

[0003] In order to solve the problems of low Doppler resolution, few pulse echo signal dimensions and few extractable features of traditional radar, a series of new arrays are used in the detection system to provide hardware support for refined feature extraction and analysis of targets. Array radars such as digital array pan-probe radars transmit omnidirectional or ultra-wide beams, without the need for beam scanning and complex resource scheduling, to achieve full space-time coverage, and through digital beamforming, weighted summation of the received signals of each array element is performed to "guide" the antenna array beam in a certain direction, forming multiple narrow beams to cover the monitoring area, and can simultaneously maintain a gaze state at all directions in the entire observation area.

[0004] Compared with scanning radar, digital array pan-probing radar has a high data rate and can obtain multi-dimensional measurement information. It can obtain high-resolution Doppler information of the target through pulse accumulation. In addition, the increase in radar signal dimensions can extract multi-dimensional and multi-domain features, and has strong detection capabilities for low-observable targets. In the current digital array radar signal target detection method, after obtaining the multi-element received signal, multiple beams with fixed azimuth pointing are formed through digital beamforming, and pulse compression and moving target detection are performed on each beam signal to obtain the range-Doppler map for detection. However, due to the complex frequency distribution characteristics of sea clutter, it is easy to submerge the target signal, especially in the weak target detection task at sea, and the performance of the Doppler feature detection method is limited. Summary of the invention

[0005] The purpose of the present invention is to solve the problem mentioned in the prior art that the beamforming-moving target detection method in the existing array radar target detection has insufficient detection performance for cross-range units and cross-beam moving targets under complex clutter background, and to provide an array radar target detection method based on Fourier transform of space-time graph data. The method extracts an advantageous angle from the high-dimensional complex structure data features of the GSP method, separates the target and clutter components in the echo signal, and improves the weak target detection capability under the sea clutter background.

[0006] In order to solve the above problems, the present application is implemented through the following technical solutions:

[0007] An array radar target detection method based on Fourier transform of space-time image data is special in that it includes the following steps:

[0008] S1. Construct array radar signal space-time graph data:

[0009] Determine a detection sample according to the received radar signal, determine the nodes in the space-time graph data in the detection sample, construct a pulse factor graph and an array element factor graph according to the nodes, and express the array radar signal space-time graph data in the form of a Cartesian product graph;

[0010] S2. Perform Fourier transform on the space-time graph data constructed in step S1:

[0011] S3. Obtain ED characteristics of the target signal:

[0012] S4. Obtain the maximum eigenvalue distance-Doppler map of the target signal:

[0013] S5. Perform target detection on the maximum eigenvalue distance-Doppler map obtained in step S4 by using an existing two-dimensional feature detection method.

[0014] Furthermore, the step S1 specifically includes the following steps:

[0015] S01. Determine the test sample:

[0016] S02. Determine the nodes in the space-time graph data:

[0017] S03. Construct impulse factor graph:

[0018] S04. Construct array element factor graph:

[0019] S05. Construct array radar signal space-time graph data:

[0020] Furthermore, the step S01 is specifically as follows:

[0021] For A digital array radar with multiple receiving elements. The signal received within a pulse repetition period is , where is the number of distance units, and the signal in each distance unit As a test sample, a total of R test samples are obtained;

[0022] Furthermore, the step S02 is specifically as follows:

[0023] Each detection sample signal Corresponding to an array signal space-time graph data, the matrix Each element in is defined as a node in the space-time graph data;

[0024] matrix is a node set containing nodes, corresponding to the detection sample signal The value of each data unit in; V r = [ ( 0 , 0 ) , ( 0 , 1 ) , ⋯ , ( P − 1 , M − 1 ) ] ;

[0025] Furthermore, the step S03 is specifically as follows:

[0026] represents the time-to-pulse factor graph, represents a node in the pulse factor graph, and each pulse echo signal corresponds to a node in the pulse factor graph;

[0027] for The edge set of express: ;

[0028] Furthermore, the step S04 is specifically as follows:

[0029] represents the array element factor graph, is a node in the array element factor graph, that is, each array element corresponds to a node in the array element factor graph;

[0030] for The edge set of It means that the covariance of each array element signal is defined as the edge weight between each array element node, that is, the adjacency matrix of the array element graph is the covariance matrix of each array element signal ;

[0031] Furthermore, the step S05 is specifically as follows:

[0032] The results obtained in step S03 and step S04 are , It is expressed as the space-time graph data of array radar signal in the form of Cartesian product graph. , including graph structure And the signal value on the graph :

[0033] , and They are and Edge set, through the adjacency matrix and express.

[0034] Furthermore, the step S2 is specifically as follows:

[0035] For distance units Radar array signal space-time data , and its space-time graph data Fourier transform is:

[0036] , expressed in matrix form as: ,

[0037] Each row corresponds to the array element factor graph A basis function of , and corresponding to an eigenvalue, rearrange the rows according to the eigenvalue size, the larger The rows corresponding to the eigenvalues ​​correspond to target signal, other The smaller eigenvalues ​​correspond to the noise signal;

[0038] Each column corresponds to the impulse factor graph A basis function of , and corresponding to an eigenvalue, according to the eigenvalue phase The columns are reordered to obtain the Doppler spectrum of each signal component.

[0039] Furthermore, the matrix of the inverse transformation of the space-time graph after Fourier transformation in step S2 is expressed as: ;

[0040] Element Factor Graph The adjacency matrix of is equivalent to the subspace decomposition of the target signal and the noise signal;

[0041] Impulse Factor Graph The factor graph of is a cyclic graph, and its adjacency matrix With characteristic vector , at this time, the adjacency matrix of the pulse dimension is equivalent to the discrete Fourier transform in computational form.

[0042] Furthermore, the step S3 is specifically as follows:

[0043] Transform the matrix It is called the ED feature of the radar array signal, and the matrix Each row corresponds to the Doppler signature of a signal component;

[0044] The row corresponding to the maximum eigenvalue in the ED feature represents the Doppler feature of the target corresponding signal.

[0045] Furthermore, the step S4 is specifically as follows:

[0046] For all the ED features of the range signal, the Doppler domain feature corresponding to the maximum eigenvalue is taken out to obtain the maximum eigenvalue distance-Doppler map :

[0047] , where for No. row, when the distance unit When there is a target in the Doppler characteristic corresponding to the maximum eigenvalue is the Doppler characteristic of the target signal;

[0048] The eigenvalue size corresponding to the Doppler feature Non-uniform linear arrangement.

[0049] This application proposes an array radar signal graph representation method for space-time graph data from the perspective of the GSP method's advantage in extracting features from high-dimensional complex structure data, and on this basis proposes a space-time graph Fourier transform target detection method to separate the target and clutter components in the echo signal, enhance the target characteristics by generating eigenvalue-Doppler features, and improve the weak target detection capability under the background of sea clutter.

[0050] Compared with the beamforming-moving target detection method, the proposed detection method is not affected by the target orientation and has more stable detection performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a flow chart of the method of the present invention;

[0052] Figure 2 It is a schematic diagram of data changes of the present invention;

[0053] Figure 3 is the ED feature map;

[0054] Figure 4 This is an example diagram of the Doppler characteristics of the target signal in the ED characteristics;

[0055] Figure 5 Schematic diagram of target detection using the DBF-MTD method;

[0056] Figure 6 Schematic diagram of target detection of this method.

[0057] illustrate: Figure 5 The target signal-to-noise ratio in is 14.04dB. Figure 6 The target signal-to-noise ratio is 38.14dB. DETAILED DESCRIPTION

[0058] The following is a specific implementation of the present invention with reference to the accompanying drawings to further illustrate the composition of the present invention.

[0059] Embodiment 1. A method for detecting target by array radar based on Fourier transform of space-time image data, the process of which is as follows: Figure 1 As shown, the data changes as Figure 2 As shown, the specific steps include:

[0060] S1. Construct array radar signal space-time graph data:

[0061] Determine a detection sample according to the received radar signal, determine the nodes in the space-time graph data in the detection sample, construct a pulse factor graph and an array element factor graph according to the nodes, and express the array radar signal space-time graph data in the form of a Cartesian product graph;

[0062] S2. Perform Fourier transform on the space-time graph data constructed in step S1:

[0063] For distance units Radar array signal space-time data , and its space-time graph data Fourier transform is: ,

[0064] It can be expressed in matrix form as: ,

[0065] Each row corresponds to the array element factor graph A basis function of , and corresponding to an eigenvalue, rearrange the rows according to the eigenvalue size, the larger The rows corresponding to the eigenvalues ​​correspond to target signal, other The smaller eigenvalues ​​correspond to the noise signal;

[0066] Each column corresponds to the impulse factor graph A basis function of , and corresponding to an eigenvalue, according to the eigenvalue phase The columns are reordered to obtain the Doppler spectrum of each signal component;

[0067] S3. Obtain ED characteristics of the target signal:

[0068] Transform the matrix It is called the ED feature of the radar array signal, as follows Figure 3 As shown, the matrix Each row corresponds to the Doppler signature of a signal component;

[0069] The row corresponding to the maximum eigenvalue in the ED feature represents the Doppler feature of the target corresponding signal, for example Figure 4 As shown;

[0070] S4. Obtain the maximum eigenvalue distance-Doppler map of the target signal:

[0071] For all the ED features of the range signal, the Doppler domain feature corresponding to the maximum eigenvalue is taken out to obtain the maximum eigenvalue distance-Doppler map :

[0072] , where for No. row, when the distance unit When there is a target in the Doppler characteristic corresponding to the maximum eigenvalue is the Doppler characteristic of the target signal;

[0073] S5, performing target detection on the maximum eigenvalue distance-Doppler map obtained in step S4 by using an existing two-dimensional feature detection method;

[0074] The two-dimensional feature detection method used in this embodiment is two-dimensional constant false alarm detection.

[0075] Furthermore, the step S1 specifically includes the following steps:

[0076] S01. Determine the test sample:

[0077] For A digital array radar with multiple receiving elements. The signal received within a pulse repetition period is , where is the number of distance units, and the signal in each distance unit As a test sample, a total of R test samples are obtained;

[0078] S02. Determine the nodes in the space-time graph data:

[0079] Each detection sample signal Corresponding to an array signal space-time graph data, the matrix Each element in is defined as a node in the space-time graph data;

[0080] matrix is a node set containing nodes, corresponding to the detection sample signal The value of each data unit in; V r = [ ( 0 , 0 ) , ( 0 , 1 ) , ⋯ , ( P − 1 , M − 1 ) ] ;

[0081] S03. Construct impulse factor graph:

[0082] represents the time-to-pulse factor graph, represents a node in the pulse factor graph, and each pulse echo signal corresponds to a node in the pulse factor graph;

[0083] for The edge set of express: ;

[0084] S04. Construct array element factor graph:

[0085] represents the array element factor graph, is a node in the array element factor graph, that is, each array element corresponds to a node in the array element factor graph;

[0086] for The edge set of It means that the covariance of each array element signal is defined as the edge weight between each array element node, that is, the adjacency matrix of the array element graph is the covariance matrix of each array element signal ;

[0087] S05. Construct array radar signal space-time graph data:

[0088] The results obtained in step S03 and step S04 are , It is expressed as the space-time graph data of array radar signal in the form of Cartesian product graph. , including graph structure And the signal value on the graph :

[0089] , and They are and Edge set, through the adjacency matrix and express.

[0090] Furthermore, the matrix of the inverse transformation of the space-time graph after Fourier transformation in step S2 is expressed as: ;

[0091] Element Factor Graph The adjacency matrix of is equivalent to the subspace decomposition of the target signal and the noise signal;

[0092] Impulse Factor Graph The factor graph of is a cyclic graph, and its adjacency matrix With characteristic vector , at this time, the adjacency matrix of the pulse dimension is equivalent to the discrete Fourier transform in computational form.

[0093] Furthermore, in step S4, the eigenvalue size corresponding to the Doppler feature Non-uniform linear arrangement.

[0094] Compared with the beamforming-moving target detection DBF-MTD method, the proposed detection method is not affected by the target orientation and has a more stable detection performance. Compared with the DBF-MTD method commonly used in current array radar target detection, the proposed method can effectively improve the weak target detection capability, such as Figure 5-6 shown.

[0095] Embodiment 2. A method for detecting a target by an array radar based on Fourier transform of space-time image data, comprising the following steps:

[0096] S1. Construct array radar signal space-time graph data:

[0097] Determine the detection sample according to the received radar signal, determine the node in the space-time graph data in the detection sample, construct the pulse factor graph and array element factor graph according to the node, and express the array radar signal space-time graph data in the form of Cartesian product graph. The specific steps are as follows:

[0098] S01. Determine the test sample:

[0099] For a digital array radar with 64 receiving elements, the 1000 range unit signals received within 512 pulse repetition periods are , the signal in each distance unit As a test sample, a total of 1000 test samples were obtained;

[0100] S02. Determine the nodes in the space-time graph data:

[0101] Each detection sample signal Corresponding to an array signal space-time graph data, the matrix Each element in is defined as a node in the space-time graph data;

[0102] matrix is a node set containing nodes, corresponding to the detection sample signal The value of each data unit in; ;

[0103] S03. Construct impulse factor graph:

[0104] represents the time-to-pulse factor graph, represents a node in the pulse factor graph, and each pulse echo signal corresponds to a node in the pulse factor graph;

[0105] for The edge set of express: ;

[0106] S04. Construct array element factor graph:

[0107] represents the array element factor graph, is a node in the array element factor graph, that is, each array element corresponds to a node in the array element factor graph;

[0108] for The edge set of It means that the covariance of each array element signal is defined as the edge weight between each array element node, that is, the adjacency matrix of the array element graph is the covariance matrix of each array element signal ;

[0109] S05. Construct array radar signal space-time graph data:

[0110] The results obtained in step S03 and step S04 are , It is expressed as the space-time graph data of array radar signal in the form of Cartesian product graph. , including graph structure And the signal value on the graph :

[0111] , and They are and Edge set, through the adjacency matrix and express;

[0112] S2. Perform Fourier transform on the space-time graph data constructed in step S1:

[0113] For distance units Radar array signal space-time data , and its space-time graph data Fourier transform is: ,

[0114] It can be expressed in matrix form as: , and its inverse transformation matrix is ​​expressed as: , array element factor graph The adjacency matrix of is equivalent to the subspace decomposition of the target signal and the noise signal;

[0115] Each row corresponds to the array element factor graph A basis function of , and corresponding to an eigenvalue, rearrange the rows according to the eigenvalue size, the larger The rows corresponding to the eigenvalues ​​correspond to target signal, other The smaller eigenvalues ​​correspond to the noise signal;

[0116] Impulse Factor Graph The factor graph of is a cyclic graph, and its adjacency matrix With characteristic vector , at this time, the adjacency matrix of the pulse dimension is equivalent to the discrete Fourier transform in computational form;

[0117] Each column corresponds to the impulse factor graph A basis function of , and corresponding to an eigenvalue, according to the eigenvalue phase The columns are reordered to obtain the Doppler spectrum of each signal component;

[0118] S3. Obtain ED characteristics of the target signal:

[0119] Transform the matrix It is called the ED feature of the radar array signal, and the matrix Each row corresponds to the Doppler signature of a signal component;

[0120] The row corresponding to the maximum eigenvalue in the ED feature represents the Doppler feature of the target corresponding signal;

[0121] S4. Obtain the maximum eigenvalue distance-Doppler map of the target signal:

[0122] For all the ED features of the range signal, the Doppler domain feature corresponding to the maximum eigenvalue is taken out to obtain the maximum eigenvalue distance-Doppler map :

[0123] , where for No. row, when the distance unit When there is a target in the Doppler characteristic corresponding to the maximum eigenvalue is the Doppler characteristic of the target signal;

[0124] S5. Perform target detection on the maximum eigenvalue distance-Doppler map obtained in step S4 by using an existing two-dimensional feature detection method.

[0125] The two-dimensional feature detection method used in this embodiment is two-dimensional constant false alarm detection.

[0126] The eigenvalue size corresponding to the Doppler feature Non-uniform linear arrangement.

[0127] The specific embodiments described herein are merely examples of the spirit of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in similar ways, but they will not deviate from the spirit of the present invention or exceed the defined scope.

Claims

1. A method for array radar target detection based on Fourier transform of space-time image data, Features: The following steps are involved: S1. Construct array radar signal space-time graph data: Determine a detection sample according to the received radar signal, determine the nodes in the space-time graph data in the detection sample, construct a pulse factor graph and an array element factor graph according to the nodes, and express the array radar signal space-time graph data in the form of a Cartesian product graph; specifically include the following steps: S01. Determine the test sample: A digital array radar with multiple receiving elements. The signal received within a pulse repetition period is , where is the number of distance units, and the signal in each distance unit As a test sample, a total of R test samples are obtained; S02. Determine the nodes in the space-time graph data: S03. Construct impulse factor graph: S04. Construct array element factor graph: S05. Construct array radar signal space-time graph data: S2. Perform Fourier transform on the space-time image data constructed in step S1: Radar array signal space-time data , and its space-time graph data Fourier transform is: ; expressed in matrix form as: , Each row corresponds to the array element factor graph A basis function of , and corresponding to an eigenvalue, rearrange the rows according to the eigenvalue size, the larger The rows corresponding to the eigenvalues ​​correspond to target signal, other The smaller eigenvalues ​​correspond to the noise signal; Each column corresponds to the impulse factor graph A basis function of , and corresponding to an eigenvalue, according to the eigenvalue phase The columns are reordered to obtain the Doppler spectrum of each signal component; S3. Obtain the ED features of the target signal: transform the matrix It is called the ED feature of the radar array signal, and the matrix Each row corresponds to the Doppler signature of a signal component; The row corresponding to the maximum eigenvalue in the ED feature represents the Doppler feature of the target corresponding signal; S4. Obtain the maximum eigenvalue distance-Doppler map of the target signal: S5. Perform target detection on the maximum eigenvalue distance-Doppler map obtained in step S4 by using an existing two-dimensional feature detection method.

2. The array radar target detection method based on Fourier transform of space-time image data according to claim 1, Features: The step S02 is specifically as follows: Each detection sample signal Corresponding to an array signal space-time graph data, the matrix Each element in is defined as a node in the space-time graph data; matrix is a node set containing nodes, corresponding to the detection sample signal The value of each data unit in; .

3. The array radar target detection method based on Fourier transform of space-time image data according to claim 1 or 2, Features: The step S03 is specifically as follows: represents the time-to-pulse factor graph, represents a node in the pulse factor graph, and each pulse echo signal corresponds to a node in the pulse factor graph; for The edge set of express: .

4. The array radar target detection method based on Fourier transform of space-time image data according to claim 3, Features: The step S04 is specifically as follows: represents the array element factor graph, is a node in the array element factor graph, that is, each array element corresponds to a node in the array element factor graph; for The edge set of It means that the covariance of each array element signal is defined as the edge weight between each array element node, that is, the adjacency matrix of the array element graph is the covariance matrix of each array element signal .

5. The array radar target detection method based on Fourier transform of space-time image data according to claim 4, Features: The step S05 is specifically as follows: The results obtained in step S03 and step S04 are , It is expressed as the space-time graph data of array radar signal in the form of Cartesian product graph. , including graph structure And the signal value on the graph : , and They are and Edge set, through the adjacency matrix and express.

6. The array radar target detection method based on Fourier transform of space-time image data according to claim 1 or 2, Features: The step S4 is specifically as follows: For all the ED features of the distance signal, the Doppler feature corresponding to the maximum eigenvalue is taken to obtain the maximum eigenvalue distance-Doppler map : , In the formula, for No. row, when the distance unit When there is a target in the Doppler characteristic corresponding to the maximum eigenvalue is the Doppler characteristic of the target signal.